See also: seed topics.
Idea for a very simple, boring language CGI/site thing: do statistical analyses of words' positions within certain great texts, and then relay the results in a "word (defn.) (ety.) (top placements to the left) (top contextual substitutions) (top placements to the right)" format. You could have a web crawler giving results as well as using dictionaries, Shakespeare, etc. From IRC, explaining it a bit more, if not much clearlyier: "I just had an idea for a boring languagey CGI. auto-generate synonyms by checking a word's context within great bits of literature, and then expose the top N results in the CGI. You could give the defintion and etymology at the top, then the most common words that appear to the left of this word in sentences, the most common ones that appear to the right, and the most common ones that are used in its stead with the same L and R."
Given that English is the lingua franca of the world, the term "lingua franca" is becoming increasingly ironic.
Sean B. Palmer